1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
|
\name{sparsesvd}
\alias{sparsesvd}
\title{Singular Value Decomposition of a Sparse Matrix.}
\usage{
sparsesvd(M, rank=0L, tol=1e-15, kappa=1e-6)
}
\arguments{
\item{M}{a sparse real matrix in \bold{Matrix} package format.
The preferred format is a \code{\link[=dgCMatrix-class]{dgCMatrix}} and other storage formats will automatically be converted if possible.
}
\item{rank}{an integer specifying the desired number of singular components, i.e. the rank of the truncated SVD.
Specify 0 to return all singular values of magnitude larger than \code{tol} (default).
}
\item{tol}{exclude singular values whose magnitude is smaller than \code{tol}}
\item{kappa}{accuracy parameter \eqn{\kappa} of the SVD algorithm (with SVDLIBC default)}
}
\value{
The truncated SVD decomposition
\deqn{
M_r = U_r D V_r^T
}
where \eqn{M_r} is the optimal rank \eqn{r} approximation of \eqn{M}.
Note that \eqn{r} may be smaller than the requested number \code{rank} of singular components.
The returned value is a list with components
\item{d}{
a vector containing the first \eqn{r} singular values of \code{M}
}
\item{u}{
a column matrix of the first \eqn{r} left singular vectors of \code{M}
}
\item{v}{
a column matrix of the first \eqn{r} right singular vectors of \code{M}
}
}
\description{
Compute the (usually truncated) singular value decomposition (SVD) of a sparse real matrix.
This function is a shallow wrapper around the SVDLIBC implementation of Berry's (1992) single Lanczos algorithm.
}
\references{
The SVDLIBC homepage \code{http://tedlab.mit.edu/~dr/SVDLIBC/} seems to be no longer available.
A copy of the source code can be obtained from \url{https://github.com/lucasmaystre/svdlibc}.
Berry, Michael~W. (1992). Large scale sparse singular value computations.
\emph{International Journal of Supercomputer Applications}, \bold{6}, 13--49.
}
\seealso{
\code{\link{svd}}, \code{\link{sparseMatrix}}
}
\examples{
M <- rbind(
c(20, 10, 15, 0, 2),
c(10, 5, 8, 1, 0),
c( 0, 1, 2, 6, 3),
c( 1, 0, 0, 10, 5))
M <- Matrix::Matrix(M, sparse=TRUE)
print(M)
res <- sparsesvd(M, rank=2L) # compute first 2 singular components
print(res, digits=3)
M2 <- res$u \%*\% diag(res$d) \%*\% t(res$v) # rank-2 approximation
print(M2, digits=1)
print(as.matrix(M) - M2, digits=2) # approximation error
}
|
